Real time decompression and post-decompress manipulation of compressed full motion video

• US 5,781,184 A
• Filed: 09/23/1994
• Issued: 07/14/1998
• Est. Priority Date: 09/23/1994
• Status: Expired due to Term

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Cagent Technologies Incorporated
• Inventors: Wallace, Gregory K., Williams, Gregory G., Wasserman, Steve C., Cassareau, Philippe M., Pansegrau, Curtis, Bucher, Timothy P.
• Primary Examiner(s): Weldon, Ulysses
• Application Number:
  US08/311,192
• Time in Patent Office: 1,390 Days
• Field of Search: 348/441, 348/443, 382/166, 382/232, 382/233, 345/202
• US Class Current: 348/571
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 5/262 Studio circuits, e.g. for mixing, switching-over, change of character of image, other special effects ; Cameras specially adapted for the electronic generation of special effects

Abstract

A method and apparatus are provided for real time decompression and post-decompress manipulation of compressed full motion video. A decompressed play-back signal is supplied to a real-time video processing system by way of an adaptor for post-decompress manipulation. The post-decompress manipulations include real-time, interactive geometry transformations and color re-mappings. The adaptor includes a range expansion module for eliminating dead space such as found in CCIR-601 outputs. The adaptor further includes a resampling module for programmably converting the pixels-per-line format of the decompressed play-back signal. The adaptor additionally includes a clip-and-dither module for altering the bits-per-pixel format of the adaptor output signal and for applying a programmably-defined dither to the adaptor output signal.

Claims

1. A full motion video system comprising:

(a) decompressor means for decompressing a supplied pre-compressed digital representation of a moving image portion of a full motion picture, where the decompressor means outputs in real time, a decompressed first video signal having a first format, the decompressed first video signal being representative of at least a first moving image portion of the full motion picture; and

(b) real-time processor means for receiving and digitally processing in real time, the first video signal or another video signal derived from the first video signal, said digital processing producing a transformed video signal representative of a transformed second moving image portion derived from the first moving image portion;

wherein said real-time processor means includes;

(b.1) geometry transform means for geometrically transforming in real time the first moving image portion as represented by the decompressed first video signal or a derivative thereof to produce therefrom a geometrically transformed signal representative of the geometrically transformed first moving image portion.

2. A full motion video system according to claim 1 wherein said real-time processor means includes:

(b.2) interactive control means for responding in real time to user actuations; and

(b.1a) wherein said geometry transform means is operatively coupled to the interactive control means for geometrically transforming in real time response to said user actuations, the first image as represented by the received first video signal or a derivative thereof and for thereby defining said transformed video signal.

3. A full motion video system according to claim 1 wherein

(b.1a) said geometry transform means imparts in real time to the first moving image portion, one or more geometry transform operations including rotation.

4. A full motion video system according to claim 1 wherein said real-time processor means further includes:

(b.2) color-remapping means for remapping colors in the first moving image portion as represented by the decompressed first video signal or a derivative thereof and for thereby defining colors of the transformed video signal.

5. A full motion video system according to claim 4 wherein said color-remapping means re-maps colors within the first moving image portion by performing in real time, one or more color re-mapping operations selected from the group consisting of:

interpolation, palette substitutions, and color blending.

6. A full motion video system according to claim 4 wherein said real-time processor means further includes:

(b.3) interactive control means for responding in real time to user actuations; and

wherein said geometry transform means and color-remapping means are operatively coupled to the interactive control means for altering their respective geometry transform and color-remapping functions in real time response to user actuations of the interactive control means.

7. A full motion video system according to claim 1 wherein the processor means is configured to process image data having a second format, said second format being different from the first format;

and said full motion video system further comprises;

(c) adaptor means, interposed between the decompressor means and the processor means, for converting the decompressed first video signal in real time into a second decompressed video signal having the second format, and for presenting the second decompressed video signal to the processor means, said second decompressed video signal being substantially representative of said first moving image portion of the full motion picture.

8. A full motion video system according to claim 7 wherein said real-time processor means includes:

(b.2) interactive control means for responding in real time to user actuations; and

(b.1a) wherein said geometry transform means is operatively coupled to the interactive control means for geometrically transforming in real time response to said user actuations, the first image as represented by the received second video signal and for thereby defining a third video signal representative of the transformed image.

9. A full motion video system according to claim 7 wherein

(b.1a) said geometry transform means includes skew means for imparting skew to the first moving image either once or recursively as the first moving image is transformed into said transformed image.

10. A full motion video system according to claim 7 wherein

(b.1a) said geometry transform means includes scaling means for scaling the size of the first image either once or recursively as the first moving image is transformed into said transformed image.

11. A full motion video system according to claim 7 wherein said real time processor means includes

(b.2) color modulation means for modulating colors of the first image either once or recursively as the first moving image is transformed into said transformed image, said modulating of colors occurring at least partially in real time response to indications by the interactive control means of user actuations.

12. A full motion video system according to claim 7 wherein the first format has a programmably-defined first arrangement of pixels per frame and the second format has a second arrangement of pixels per frame and the adaptor means is programmable to accommodate said programmably-defined first arrangement of pixels per frame.

13. A full motion video system according to claim 12 wherein:

the programmably;

defined first arrangement of pixels per frame varies;

the second arrangement of pixels per frame is programmably-defined to vary in correspondence with variations of the first arrangement; and

the adaptor means includes a programmable resampling means for producing, in response to variations of the first arrangement, the corresponding second arrangement of pixels per frame.

14. A full motion video system according to claim 13 wherein:

the programmably-defined first arrangement has either 352-by-240 or 352-by-288 pixels-per-frame and the corresponding second arrangement respectively has either 320 pixels-per-line or 384 pixels-per-line.

15. A full motion video system according to any one of claims 7, 9, 10, 12 and 13 wherein:

the second format has symmetric pixels each of equal height and width; and

the first format has asymmetric pixels each having a width that is different from a respective height of the respective asymmetric pixel.

16. A full motion video system according to claim 7 wherein the first format has a predefined first number of pixels-per-line and the second format has a different second number of pixels-per-line.

17. A full motion video system according to claim 16 wherein the second number of pixels-per-line is programmably-defined and the adaptor means includes:

(c.1) programmable resampling means for producing, in response to a supplied programming signal, the second format with the corresponding second number of pixels-per-line.

18. A full motion video system according to claim 17 wherein said programmable resampling means includes:

(c.11) sequencer means responsive to said programming signal; and

(c.12) weight and sum means, responsive to the sequencer means, for receiving a source signal representing sequential source pixels, for assigning respective weights to the source pixels, and for producing a weighted sum signal representing a sum of the respectively weighted source pixels.

19. A full motion video system according to claim 18 wherein said weight and sum means comprises:

(c.121) first multiples producing means, responsive to the sequencer means, for receiving the source signal and for producing a first multiple signal representing a sequencer-selected one of a constant and a first set of multiples of the sequential source pixels;

(c.122) second multiples producing means, responsive to the sequencer means, for receiving the source signal and for producing a second multiple signal representing a sequencer-selected one of a constant and a second set of multiples of the sequential source pixels; and

(c.123) third multiples producing means, responsive to the sequencer means, for receiving the source signal and for producing a third multiple signal representing a sequencer-selected one of a constant and a third set of multiples of the sequential source pixels; and

(c.124) summing means, operatively coupled to the first, second and third multiples producing means, for producing a summation signal representing a sum of the first through third multiple signals.

20. A full motion video system according to claim 19 wherein said weight and sum means further comprises:

(c.125) first register means, coupled to the summing means, for storing a first instance of the summation signal;

(c.126) second register means, coupled to the summing means, for storing a second instance of the summation signal; and

(c.127) adding means, coupled to the first and second register means, for adding the first and second instances of the summation signal and thereby producing the weighted sum signal.

21. A full motion video system according to claim 18 wherein said programmable resampling means further includes:

(c.13) programmable divider means, operatively coupled to the weight and sum means, for producing a divided signal representing the value of the weighted sum signal divided by a programmably-defined divisor.

22. A full motion video system according to claim 21 wherein the programmably-defined divisor is approximately equal to a value selected from the group consisting of 10 and 12.

23. A full motion video system according to claim 21 wherein the programmably-defined divisor is a value, Y, defined by the equation:

space="preserve" listing-type="equation">1/Y=(W+1)·

(1/X)+(W+1)·

(1/(U·

X)),

where U, W and X are programmably-defined nonzero integers each equal to two raised to a positive power.

24. A full motion video system according to claim 23 wherein the integer W is programmably selected from the group consisting of 2 and 4 and the integer X is programmably selected from the group consisting of 32 and 64.

25. A full motion video system according to claim 23 wherein the programmable divider means comprises:

(c.131) first and second left shifting means, operatively coupled to receive the weighted sum signal and to produce therefrom respective first and second shifted sum signals representing the value of the weighted sum signal shifted left by a predefined first and second number of bit positions;

(c.132) first selector means for programmably selecting and outputting one of the first and second shifted sum signals, the output of said first selector means representing the value, S times W, where S is the value of the weighted sum signal;

(c.133) first addition means, operatively coupled to receive the output of the first selector means and to receive the weighted sum signal and to produce therefrom a first addition signal representing the value, S·

(W+1);

(c.134) first right shifting means, operatively coupled to receive the first addition signal, for shifting the first addition signal right by a third predefined number of bit positions to produce therefrom a right-shifted first signal representing the value, S·

(W+1)/U;

(c.135) second addition means, operatively coupled to receive the output of the first addition means and of the first right shifting means, to produce therefrom a second addition signal representing the value, S·

(W+1)·

(1+1/U);

(c.136) second and third right shifting means, operatively coupled to receive the second addition signal and to produce therefrom respective second and third right-shifted signals representing the value of the second addition signal shifted right respectively by a predefined fourth and fifth number of bit positions; and

(c.137) second selector means for programmably selecting and outputting one of the second and third right-shifted signals, the output of said second selector means representing the value, S·

(W+1)·

(1+1/U)·

(1/X).

26. A full motion video system according to claim 7 wherein the first format has a predefined first number of bits-per-pixel and the second format has a different second number of bits-per-pixel.

27. A full motion video system according to claim 26 wherein the second number of bits-per-pixel is smaller than the first number of bits-per-pixel and wherein the adaptor means includes:

(c.1) truncation means for removing excessive, less significant bits from a precursor signal having the first number of bits-per-pixel and thereby producing a BPP-reduced signal having the smaller second number of bits-per-pixel; and

(c.2) pre-truncation dither means for adding a dither signal to a less significant portion of the precursor signal before the precursor signal is truncated by the truncation means.

28. A full motion video system according to claim 27 wherein the dither means comprises:

(c.21) adder means having a first input port for receiving the precursor signal, a second input port for receiving the dither signal, and an output port for outputting at least the non-truncated portion of the BPP-reduced signal; and

(c.22) programmable storage means, coupled to the second input port of the adder means, for storing and supplying a series of predefined dither values to the second input port to thereby define the dither signal.

29. A full motion video system according to claim 28 wherein the dither means further comprises:

(c.23) programmable sequencer means, operatively coupled to the programmable storage means and responsive to a supplied sequence-defining signal, for sequencing through the dither values that are stored in the storage means in accordance with a sequence defined by the supplied sequence-defining signal.

30. A full motion video system according to claim 29 wherein the programmable sequencer means comprises:

(c.231) a first signal generator for generating a first addressing signal representing a relatively random series of addresses;

(c.232) a second signal generator for generating a second addressing signal representing a non-random, ordered series of addresses; and

(c.233) selection means, responsive to the sequence-defining signal and operatively coupled to the first and second signal generators, for selecting one of the first and second addressing signals as a storage addressing signal that defines said sequencing through the dither values stored in the storage means.

31. A full motion video system according to claim 27 wherein:

the first format has a plurality of first color components each having a first number of bits-per-component;

the second format has a plurality of second color components each having a second, smaller number of bits-per-component;

the truncation means comprises a plurality of independent truncating sections each corresponding to a respective one of the plural second color components, each truncating section removing excessive, less significant bits from a corresponding color component of the precursor signal to thereby produce the BPP-reduced signal with the smaller second number of bits-per-component; and

the dither means comprises a plurality of independent dithering sections each corresponding to a respective one of the plural second color components and coupled to a respective one of the truncating sections, each dithering section adding a respective, component-dither signal to a less significant portion of the corresponding color component in the precursor signal before the precursor component is truncated by a corresponding one of the truncating sections.

32. A full motion video system according to claim 31 wherein:

the first format has at least three first color components each having at least eight bits-per-component;

the second format has a corresponding at least three second color components each having less than eight bits-per-component; and

the respective component-dither signal of each dithering section has a number of bits at least equal to the difference in number of bits per component between the corresponding first and second color components.

33. A full motion video system according to claim 31 wherein:

each of the independent dithering sections includes a respectively independent, adder means having a first input port for receiving the corresponding color component of the precursor signal, a second input port for receiving the corresponding component dither signal, and an output port for outputting at least the non-truncated portion of the corresponding color component in the BPP-reduced signal; and

each of the independent dithering sections further includes a respectively independent, programmable storage means, coupled to the second input port of the corresponding adder means, for storing and supplying a series of predefined dither values to the corresponding second input port to thereby independently define the corresponding component dither signal.

34. A full motion video system according to claim 33 wherein:

each of the independent dithering sections additionally includes a respectively independent, programmable sequencer means, operatively coupled to the corresponding programmable storage means and responsive to a correspondingly supplied, component sequence-defining signal, for sequencing through the dither values that are stored in the corresponding storage means in accordance with a sequence defined by the correspondingly supplied, component sequence-defining signal.

35. A full motion video system according to claim 7 wherein the first format has a predefined first number of bits-per-pixel and the decompressed first video signal uses a first quantization range that is less than the maximum quantization range possible with said first number of bits-per-pixel, the unused portion or portions of the first quantization range defining a corresponding one or more dead spaces, and further wherein the adaptor means comprises:

(c.1) range expansion means for receiving the decompressed first video signal and producing therefrom a range-expanded signal that uses a second quantization range that is wider than the first quantization range.

36. A full motion video system according to claim 35 wherein the range expansion means performs range expansion so as to approximately satisfy the following expansion formula:

space="preserve" listing-type="equation">R'"'"'=(R-L)·

(U'"'"'-L'"'"'+1)/(U-L+1)

where R represents a component value of the decompressed first video signal, R'"'"' represents the corresponding expanded component value, L represents a predefined lower bound of the range-limited R value, U represents a predefined upper bound of the range-limited R value, L'"'"' represents a predefined lower bound of the expanded R'"'"' value, and U'"'"' represents a predefined upper bound of the expanded R'"'"' value.

37. A full motion video system according to claim 36 wherein the range expansion means includes approximate multiplying means for multiplying a difference signal D representing (R-L) by an approximation of the multiplying factor, (U'"'"'-L'"'"'+1)/(U-L+1), said approximation being equal to a sum of three or less terms, each term being equal to an integer multiple of the value 2 raised to an integer power.

38. A full motion video system according to claim 36 wherein the range expansion means includes approximate multiplying means for multiplying a difference signal representing D=(R-L) by an approximation of the multiplying factor, (U'"'"'-L'"'"'+1)/(U-L+1), said approximate multiplying means comprising;

(c.1) first left-shifting means for receiving the difference signal and producing therefrom a first left-shifted difference signal representing D multiplied by 2 raised to a first predefined power, the first predefined power being a positive first integer;

(c.2) first adder means having a first input for receiving the difference signal and a second input for receiving the first left-shifted difference signal, said first adder means producing a first sum signal representing the sum of the difference signal and the first left-shifted difference signal;

(c.3) second left-shifting means for receiving the first sum signal and producing therefrom a second left-shifted difference signal representing the value of the first sum signal multiplied by 2 raised to a second predefined power, the second predefined power being a positive second integer;

(c.4) second adder means having a first input for receiving the difference signal and a second input for receiving the second left-shifted difference signal, said second adder means producing a second sum signal representing the sum of the difference signal and the second left-shifted difference signal.

39. A full motion video system according to claim 36 wherein the approximate multiplying means further comprises:

(c.5) right-shifting means for receiving the second sum signal and producing therefrom a justified signal representing the value of the second sum signal multiplied by 2 raised to a second predefined power, the second predefined power being a negative third integer.

40. A full motion video system according to claim 1 wherein the first format has a predefined first number of bits-per-pixel, the decompressed first video signal uses a first quantization range that is less than the maximum quantization range possible with said first number of bits-per-pixel, the unused portion or portions of the first quantization range defining a corresponding one or more dead spaces, the color of each pixel of the pre-compressed digital representation is defined in a first color space, the color of each pixel of the decompressed first video signal is defined in a different second color space, and further wherein the decompressor means comprises:

(a.1) color space conversion means for converting color signals representing colors according to coordinates of the first color space into converted color signals representing colors according to coordinates of the second color space in accordance with the following linear transformation, ##EQU4## wherein R, G, and B are coordinate values of the second color space, Y, C_R and C_B are coordinate values of the first color space, g is a gain factor that is appropriately set to eliminate or reduce an upper dead space of the first quantization range, and off is an offset factor that may be set to zero or appropriately set to eliminate or reduce a lower dead space of the first quantization range.

41. A full motion video system according to claim 1 wherein the pre-compressed digital representation is stored on a disk and said system further comprises:

(c) a disk playback mechanism, operatively coupled to the decompressor means, for reproducing the pre-compressed digital representation on a real time basis and supplying the reproduced digital representation to the decompressor means for real time decompression and subsequent geometric transformation by said geometry transform means.

42. A full motion video system according to claim 41 wherein the disk includes optical data storage areas and the disk playback mechanism includes optical playback means for detecting data optically stored on the optical data storage areas and for reproducing such optically stored data.

43. A full motion video system according to claim 41 wherein the disk stores both compressed video and audio portions of a full motion picture and the decompressor means outputs in real time, a decompressed audio signal representative of the audio portion of the full motion picture.

44. A full motion video system according to claim 1 wherein:

(a.1) the decompressor means is further for decompressing a pre-compressed digital representation of a audio portion of a full motion picture, and where the decompressor means outputs in real time, a decompressed audio signal representative of at least a first audio portion of the full motion picture; and

(b.1) the real-time processor means is further for digitally processing in real time, the first audio portion represented by the decompressed audio signal to produce therefrom a transformed audio signal representative of a transformed audio portion derived from the first audio portion.

45. A full motion video system according to claim 3 wherein:

(b.1b) said geometry transform operations further include skewing.

46. A full motion video system according to claim 45 wherein:

(b.1c) said geometry transform operations further include size rescaling.

47. A full motion video system according to claim 3 wherein:

(b.1b) said geometry transform means imparts the one or more geometry transform operations recursively in real time.

48. A full motion video system according to claim 16 wherein the ratio between said first number of pixels-per-line and said second number of pixels-per-line is not a power of two.

49. A full motion video system according to claim 16 wherein said first number of pixels-per-line is at least 352.

50. A full motion video system according to claim 40 wherein said offset factor is set to a value other than zero so as to eliminate or reduce a lower dead space of the first quantization range.

51. A full motion video system according to claim 44 wherein said real-time processor means includes:

(b.1a) sound processing means for processing said decompressed audio signal to create an impression of a sound source that is moving through 3-dimensional space.

52. A machine-implemented method for producing a real-time-manipulated full motion picture comprising the steps of:

(a) retrieving from a compact disk or another storage means which has a compressed full-motion picture stored therein, compressed data representing video and audio of at least a portion of the full-motion picture;

(b) decompressing the retrieved data;

(c) transmitting the decompressed data in real-time to a real-time data processing system for real-time manipulation;

(c'"'"') performing said real-time manipulation on the transmitted data; and

(d) outputting the real-time manipulated data for appreciation by a human recipient;

wherein said step of performing the real-time manipulation includes;

geometrically transforming in real-time the video portion of the transmitted data so as to produce a geometrically transformed version of the transmitted video portion.

53. A production method according to claim 52 wherein:

(c.1) the real-time manipulation includes imparting threedimensional effects in real time to the decompressed audio portion of the transmitted data so as to produce a transformed version of the audio portion of the transmitted data.

54. A production method according to claim 52 wherein the video portion of the decompressed data includes video data having a first format, the real-time data processing system is structured to manipulate video data having a different second format, and said method further comprises, before the step of transmitting the decompressed data to the real-time data processing system, the additional step of:

(e) converting the format of the decompressed video data in real time from said first format to said second format.

55. A production method according to claim 54 wherein the first format has a first arrangement of pixels per frame, the second format has a programmably-defined second arrangement of pixels per frame, and said step of converting includes:

(e.1) automatically determining what programmably-defined arrangement of pixels per frame is being employed in the second format; and

(e.2) in response to said determination, automatically converting the video data from said first arrangement of pixels per frame said second arrangement of pixels per frame.

56. A production method according to claim 55 wherein the first arrangement of pixels per frame uses a first number of pixels per image line, the second arrangement of pixels per frame uses a different second number of pixels per image line, said step of responsively converting in real time includes:

(e.2a) resampling each image line of the decompressed video data in real time and thereby changing the number pixels per image line from said first number to said second number.

57. A production method according to claim 54 wherein the first format uses a first number of bits-per-pixel and the second format uses a smaller second number of bits-per-pixel, and said step of converting includes:

(e.1) reducing the number of bits-per-pixel from said first number to said smaller second number.

58. A production method according to claim 57 wherein said step of reducing includes:

(e.1a) adding a dither signal to the decompressed video data.

59. A production method according to claim 58 wherein said step of adding a dither signal includes:

(e.1b) programmably selecting the dither signal from the group consisting of a random dither signal and an ordered-sequence dither signal.

60. A production method according to claim 54 wherein the wherein the first format has a predefined first number of bits-per-pixel and the decompressed video signal uses a first quantization range that is less than the maximum quantization range possible with said first number of bits-per-pixel, the unused portion or portions of the first quantization range defining a corresponding one or more dead spaces, and said step of converting includes:

(e.1) expanding the quantization range so as to reduce or eliminate one or more of the dead spaces.

61. A production method according to claim 52 wherein said step of geometrically transforming includes at least one of:

(c'"'"'.1) rotating the video portion of the transmitted data, and

(c'"'"'.2) skewing the video portion of the transmitted data.

62. A production method according to claim 61 wherein said step of geometrically transforming further includes:

(c'"'"'.3) size-rescaling the video portion of the transmitted data.

63. A production method according to claim 62 wherein said step of geometrically transforming further includes:

(c'"'"'.4) recursively performing at least one of said steps of rotating, skewing and size-rescaling.

64. A production method according to claim 62 wherein said step of performing the real-time manipulation further includes:

(c'"'"'.4) performing real time color re-mapping on the video portion of the transmitted data.

65. A dithering circuit comprising:

(a) adder means having a first input port for receiving a precursor signal, a second input port for receiving a dithering signal, and an output port for outputting a dithered signal;

(b) programmable storage means, coupled to the second input port of the adder means, for storing and supplying a series of predefined dither values to the second input port to thereby define the dithering signal; and

(c) programmable sequencer means, operatively coupled to the programmable storage means and responsive to a supplied sequence-defining signal, for sequencing through the dither values that are stored in the storage means in accordance with a sequence defined by the supplied sequence-defining signal.

66. A circuit for generating signals representing weighted sums of digital input values represented by an input digital signal, said circuit comprising:

(a) first multiplexer means, coupled to receive the input digital signal and/or shifted derivatives of the input digital signal, for selectively outputting either a predefined first constant-value signal or a received one of the input digital signal and its shifted derivatives;

(b) second multiplexer means, coupled to receive the input digital signal and/or shifted derivatives of the input digital signal, for selectively outputting either a predefined second constant-value signal or a received one of the input digital signal and its shifted derivatives;

(c) first summing means for receiving the outputs of the first and second multiplexer means and for producing a first sum signal representing a sum of the outputs of the first and second multiplexer means;

(d) selection control means, operatively coupled to the first and second multiplexer means, for selecting the signal output by each of the first and second multiplexer means and thereby defining a weight to be applied to the input digital signal; and

(e) second summing means for receiving first and second outputs of the first summing means and producing therefrom a second sum signal.

67. A range-expansion circuit for receiving an input digital signal having a first number of bits but representing a set of values having a first quantization range whose extent is less than two raised to a power equal to said first number, where the difference between the first quantization range and two raised to said power defines both upper and lower deadspaces about the first quantization range, said range-expansion circuit comprising:

(a) subtraction means for subtracting from said input digital signal an offset value representing all or part of the lower deadspace and for producing therefrom an offset digital signal; and

(b) multiplying means for multiplying the offset digital signal by a gain factor to produce therefrom an expanded range signal having a second quantization range whose extent is larger than that of the first quantization range.

68. A moving picture system comprising:

(a) a real-time decompressor for decompressing a pre-compressed digital representation of an image portion of a moving picture, where the pre-compressed digital representation is supplied in real-time and the decompressor responsively outputs in real-time a decompressed first video signal representative of at least a first moving image portion of the moving picture;

(b) a reprogrammable memory, operatively coupled to the decompressor, for receiving and storing a second video signal, said second video signal being derived from the decompressed first video signal and being also representative of the at least first moving image portion of the moving picture; and

(c) a real-time image transformer operatively coupled to said memory for accessing the second video signal stored therein and for transforming the second video signal in real-time to thereby produce a transformed video signal representative of a transformed version of said at least first moving image portion of the moving picture;

wherein said real-time image transformer includes;

(c.1) geometry transform means for geometrically transforming in real time the first moving image portion as represented by the stored second video signal or a derivative thereof to produce therefrom a geometrically transformed signal representative of the transformed version of the first moving image portion.

69. A moving picture system according to claim 68 further comprising:

(d) video output means for receiving the transformed video signal and outputting in real-time to a display means, a third video signal derived from said transformed video signal.

70. A moving picture system according to claim 68 wherein

(c.1a) said geometry transform means imparts in real time to the first moving image portion as represented by the second video signal or a derivative thereof, one or more geometry transform operations including rotation.

71. A moving picture system according to claim 70 wherein

(c.1b) said geometry transform operations further include skewing.

72. A moving picture system according to claim 71 wherein

(c.1c) said geometry transform operations further include size rescaling.

73. A moving picture system according to claim 68 wherein

(c.2) said real-time image transformer returns the transformed video signal to the reprogrammable memory for storage in said memory such that the transformed video signal may be recursively transformed by the real-time image transformer.